{
  "id": 9019920,
  "title": "MRH Trowe Lets Employees Build Their Own Agent Factory",
  "url": "https://urgent.news/2026/09/21/mrh-trowe-lets-employees-build-their-own-agent-factory",
  "topic": "ai",
  "section": "AI",
  "published": "2026-09-21T21:47:36.000Z",
  "source": {
    "name": "PYMNTS",
    "slug": "pymnts",
    "url": "https://www.pymnts.com/news/artificial-intelligence/2026/mrh-trowe-lets-employees-build-their-own-agent-factory/"
  },
  "original_language": "en",
  "account": "At insurance broker MRH Trowe, employees sought assistance from artificial intelligence to streamline their daily tasks. Initially, teams experimented with AI tools independently, prompting concerns over the security of sensitive client and insurance data. To address this issue, MRH Trowe developed a secure system enabling employees to construct and utilize their own AI agents. Approximately 400 staff members gained access to this system during its initial month, according to Amazon Web Services' case study. The monthly cost for cloud and AI model services amounted to roughly $14 per employee.\n\nThe first agent, developed by the firm, efficiently compiles meeting notes. Employees communicate with the agent in German to retrieve meeting details from the calendar, extract the transcript, and generate minutes containing participant names, agenda, topics, and action items. Employees do not require extensive technical expertise to build an agent; a few lines of code suffice. MRH Trowe aims to automate repetitive work via agents designed by its subject-matter experts. Board member Leonid Karlinski encouraged colleagues to automate redundant tasks by creating LibreChat agents, which facilitate interaction with these AI agents. Two safeguards were implemented to protect client data – each agent operates under the identity of the logged-in employee, limiting access to the individual's calendar and transcripts, and all processing occurs within AWS' Frankfurt region, ensuring data remains in Germany.\n\nFinancial institutions have been early adopters of AI, with adoption rates reaching 27 out of 75 business tasks in a PYMNTS Intelligence survey of 60 senior technology executives at large U.S. companies. The sectors exhibiting high adoption are financial services (65%), credit risk assessment (60%), and sales forecasting (60%). However, revenue growth and customer-related tasks remain lagging. MRH Trowe's latest agent, \"talk to your data,\" assists with cross-selling and upselling by merging customer records with public information. A case study from BNY Mellon highlighted a contract review agent that reduced legal review time from four hours to just one, achieved through OpenAI's technology.\n\nWhile the initial case study at MRH Trowe reported user numbers and cost, figures for time saved, revenue gained, or accuracy were not disclosed. The firm plans to have 10 to 15 agents created and maintained by its subject-matter experts by the end of 2026. The financial services sector as a whole anticipates raising AI budgets by 85% within the next year, driven by the need to overcome fragmented or subpar data, which remains the most significant barrier for 30% of firms.",
  "summary": "Employees at insurance broker MRH Trowe wanted artificial intelligence to help with their daily work. Teams started trying AI tools on their own. That created a risk for the company. Sensitive client and insurance data could end up in software nobody at the firm controlled. MRH Trowe’s answer was to build one secure system where […] The post MRH Trowe Lets Employees Build Their Own Agent Factory…",
  "key_points": [],
  "editors_take": null,
  "illustration": null,
  "coverage": {
    "outlets": 1,
    "also_reported_by": []
  },
  "ai_generated": true,
  "disclaimer": "Summaries, key points and the editor’s take are written by software from other outlets’ reporting and may contain errors — always check the linked original."
}